Author Topic: ATX Bench power supply, ideas and suggestions wanted  (Read 12961 times)

0 Members and 1 Guest are viewing this topic.

Offline Ludvig82Topic starter

  • Contributor
  • Posts: 18
  • Country: no
ATX Bench power supply, ideas and suggestions wanted
« on: November 30, 2015, 10:27:35 pm »
Hi there, this is my first post on this forum.

I have wanted to build myself a bench power supply from atx power supply for a while now. I dont often need something like that, but i am alway up for learning new stuff, and i see that it may come in very handy.

In my search for this, i came over this product: http://www.seeedstudio.com/depot/atx-breakout-board-bench-power-supply-p-1222.html?cPath=155. It looks like a nice product and almost excact what i want already done for me. Does it look any good to you guys? If i buy that one, i still would like to add a variable output, but i am pretty green at making these kind of circuits, or circuits at all. I would really love if someone helped me with a suggestion how i could get this made, wich parts i need and how to connect it. To explain a bit more, i read that i can use the range between the -12v and the 12v, so the range of the output would be somewhere like 1v to 24v, depends on what parts you use. I also would love to have a display showing volts on this output so i can adjust it easily without measuring it with an external device.

Just to clear a few things up, since I foud this breakoutboard, i wont touch the inside of the PSU. I will make a case to house it all in. Not sure what the case will be yet, but that will come later :)

Hope i made myself understood ok, English is not my language, but i hope i am good enough in writing it :)

Never argue with an idiot, first he brings you down to his level. Then he beats you with experience!
 

Offline Cliff Matthews

  • Supporter
  • ****
  • Posts: 1910
  • Country: ca
    • General Repair and Support
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #1 on: December 01, 2015, 01:40:23 am »
Hello and welcome! I might as well tell you now.. That question has been asked here before and you would do well to use and read the forum searches on just that topic. Here is one such example: https://www.eevblog.com/forum/beginners/beginner-need-advice-on-atx-psu-gt-lab-psu/
I love that quote from user Sigmoid, "Converting an ATX PSU into a benchtop lab supply is - quite literally - like converting a Caterpillar hydraulic excavator into a remote surgery robot."
« Last Edit: December 01, 2015, 01:46:05 am by Cliff Matthews »
 

Offline MarkF

  • Super Contributor
  • ***
  • Posts: 2523
  • Country: us
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #2 on: December 01, 2015, 02:03:23 am »
I have the Seeedstudio ATX breakout board you mentioned.  It's okay and you can get plastic cover for the top and bottom.  However, I don't recommend it for beginners and you can NOT form a 1V to 24V output from an ATX power supply.

I would recommend a proper variable power supply that will provide you with more protection.  You can find inexpensive power supplies on eBay that would give you what you want.  Look at this HP 6216A 0-25V DC Power Supply for example.
 

Offline moya034

  • Regular Contributor
  • *
  • Posts: 84
  • Country: us
  • Former Welder, IT guy, and Ham Radio Junkie
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #3 on: December 01, 2015, 02:09:04 am »
I did this once. (i used an AT instead of ATX, but still the same bag)

I was powering a radio transceiver with it. It worked very well... until the PSU blew up and took the radio with it. I will never do it again. Life is too short to use shitty power supplies.
 



Offline moya034

  • Regular Contributor
  • *
  • Posts: 84
  • Country: us
  • Former Welder, IT guy, and Ham Radio Junkie
 

Offline Cliff Matthews

  • Supporter
  • ****
  • Posts: 1910
  • Country: ca
    • General Repair and Support
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #7 on: December 01, 2015, 12:48:19 pm »
Here's a great way to spend just $10
http://www.ebay.com/itm/Red-0-30V-2mA-3A-Continuously-Adjustable-DC-Regulated-Power-Supply-DIY-Kit-PCB-/221739263622?hash=item33a0b0a686:g:JwQAAOSwymxVJfou
You need a transformer though...  ;)
Oops, sorry! I tear so much stuff apart (because I learn and it's fun..) that I always seem to have too many hanging around.
 

Offline MarkF

  • Super Contributor
  • ***
  • Posts: 2523
  • Country: us
 

Offline Cliff Matthews

  • Supporter
  • ****
  • Posts: 1910
  • Country: ca
    • General Repair and Support
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #9 on: December 01, 2015, 02:55:41 pm »
For a few dollars more you can have one with more capabilities:

http://www.aliexpress.com/item/B3008-precision-CNC-DC-DC-constant-voltage-constant-current-buck-LED-driver-modules-solar-charging/1550355670.html


An ideal bench supply needs to have a stable and clean output (low ripple). A 24v 100va transformer with the PCB mentioned would likely prove to be the better option.
 

Offline macboy

  • Super Contributor
  • ***
  • Posts: 2252
  • Country: ca
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #10 on: December 01, 2015, 04:16:36 pm »
I am not a fan of using an ATX supply this way, but at least That breakout board includes polyfuses for some measure of protection. As you know, any ATX supply can crank out dozens of Amps on +5 or +12 V rails, which can lead to dangerous situations if you short something out. The polyfuses should at least prevent fires.

For a variable output, look for a DC-DC "buck boost" converter. These are a two-stage device, first a boost converter to increase the input to some higher intermediate voltage (maybe around 30 V), followed by a buck converter that reduces that intermediate voltage back down to the desired output voltage. These can output a wide voltage range (e.g. 2 to 30 V) from a wide input range (e.g. 3 to 30 V). If you use just a boost conveter, the output must be higher than the input, and with just a buck, the output must be lower than the input. This is an example of (but not necessarily an endorsement for) such a module: http://www.ebay.com/itm/DC-DC-Boost-Buck-Converter-Step-Up-Step-Down-Supply-Module-3-35V-to-2-2-30V-HG-/271720088700

You can then add a 4-digit voltmeter that will show (e.g.) 00.00 V to 30.00 V. Get a 3-wire one (separate input and power supply). Use the +12 rail for the power supply. This is an example (again not an endorsement): http://www.ebay.com/itm/Mini-0-33V-Car-Motorcycle-DC-4-Digital-Voltmeter-Red-LED-Voltage-Panel-Meter-New-/390734584022 . You can also get ammeter versions of these.

edit: While browsing for "DC-DC buck boost" on ebay, I see a lot of improperly labelled devices, for example, buck (step down) converters being called "buck boost" in the title. If you want a true buck-boost conveter, make sure that the photo of the device has two inductors (they usually look like the round-cornered black thingies in the one I linked to, or distinctly like coils of wire). Most of these sellers do not understand the stuff they sell, and their titles and descriptions are based on popular search terms and/or copied from other listings. Usually, in my experience, the photo is the most accurate part of these listings, so if you understand what you are looking at, then you can and should ignore much of what is written in the title and description.
« Last Edit: December 01, 2015, 04:45:04 pm by macboy »
 

Offline rdl

  • Super Contributor
  • ***
  • Posts: 3665
  • Country: us
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #11 on: December 01, 2015, 04:53:11 pm »
For someone's first bench power supply, and one that may be their only power supply for a while, a switch mode type is not a good choice. And for use as a lab power source, a modified ATX power supply is probably the worst kind of switch-mode power supply there is.
 

Offline Ludvig82Topic starter

  • Contributor
  • Posts: 18
  • Country: no
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #12 on: December 01, 2015, 06:59:45 pm »
Thansk for all aswers. I was expecting some negativity around this. I know this is not the best bench supply and that many people are against it. Mind you, let me explain my reasons for wanting one a little better. First of all it is the DIY ascpect of it, i love making things and trying new things to learn. So i figured this would be a nice project for me, and it can help me alot at work where i test alot of things with different power input. These things (Stuff for fish farming, controllers for feeding systems and so on) are most of the time not depending on stable voltage or clean power. They are often powered by car batteries or similair. So just for test purposes to see if things work, this would be ok. The stuff i am talking about is often on fish cages out at sea, and therefor often destroyed by water or similair.

Second i already have a proper bench power supply, but i would never bring that one at work on sea.

Third, i know this may be crude and so on, but still want to try it. So i still would love some more suggestions on what to do to get a adjustable voltage output, and what parts i need for it. I am not very experienced with electronice as is, but very used to power and so on. It is just when i comes to circuits and so on i am inexperienced. Thats what i want to learn. So please explain as i was a kid :P

I have seen people making tutorials where they use -12v and 12v to make adjustable output that gives more than 12v. Can someone explain this?
Never argue with an idiot, first he brings you down to his level. Then he beats you with experience!
 

Offline Ludvig82Topic starter

  • Contributor
  • Posts: 18
  • Country: no
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #13 on: December 01, 2015, 07:04:11 pm »
I am not a fan of using an ATX supply this way, but at least That breakout board includes polyfuses for some measure of protection. As you know, any ATX supply can crank out dozens of Amps on +5 or +12 V rails, which can lead to dangerous situations if you short something out. The polyfuses should at least prevent fires.

For a variable output, look for a DC-DC "buck boost" converter. These are a two-stage device, first a boost converter to increase the input to some higher intermediate voltage (maybe around 30 V), followed by a buck converter that reduces that intermediate voltage back down to the desired output voltage. These can output a wide voltage range (e.g. 2 to 30 V) from a wide input range (e.g. 3 to 30 V). If you use just a boost conveter, the output must be higher than the input, and with just a buck, the output must be lower than the input. This is an example of (but not necessarily an endorsement for) such a module: http://www.ebay.com/itm/DC-DC-Boost-Buck-Converter-Step-Up-Step-Down-Supply-Module-3-35V-to-2-2-30V-HG-/271720088700

You can then add a 4-digit voltmeter that will show (e.g.) 00.00 V to 30.00 V. Get a 3-wire one (separate input and power supply). Use the +12 rail for the power supply. This is an example (again not an endorsement): http://www.ebay.com/itm/Mini-0-33V-Car-Motorcycle-DC-4-Digital-Voltmeter-Red-LED-Voltage-Panel-Meter-New-/390734584022 . You can also get ammeter versions of these.

edit: While browsing for "DC-DC buck boost" on ebay, I see a lot of improperly labelled devices, for example, buck (step down) converters being called "buck boost" in the title. If you want a true buck-boost conveter, make sure that the photo of the device has two inductors (they usually look like the round-cornered black thingies in the one I linked to, or distinctly like coils of wire). Most of these sellers do not understand the stuff they sell, and their titles and descriptions are based on popular search terms and/or copied from other listings. Usually, in my experience, the photo is the most accurate part of these listings, so if you understand what you are looking at, then you can and should ignore much of what is written in the title and description.

Alot of interesting reading. Is it ok to pm u for more info?
Never argue with an idiot, first he brings you down to his level. Then he beats you with experience!
 

Offline Ludvig82Topic starter

  • Contributor
  • Posts: 18
  • Country: no
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #14 on: December 01, 2015, 07:05:05 pm »
Hello and welcome! I might as well tell you now.. That question has been asked here before and you would do well to use and read the forum searches on just that topic. Here is one such example: https://www.eevblog.com/forum/beginners/beginner-need-advice-on-atx-psu-gt-lab-psu/
I love that quote from user Sigmoid, "Converting an ATX PSU into a benchtop lab supply is - quite literally - like converting a Caterpillar hydraulic excavator into a remote surgery robot."

Thansk for the welcome :)
 I did a search, and already read that post. Wanted my own thread so i can ask all my silly noobquestions :)
Never argue with an idiot, first he brings you down to his level. Then he beats you with experience!
 

Offline macboy

  • Super Contributor
  • ***
  • Posts: 2252
  • Country: ca
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #15 on: December 01, 2015, 07:54:34 pm »
Alot of interesting reading. Is it ok to pm u for more info?
Post any questions here and I'll do my best to answer them for the benefit of all. That's the beauty of a public forum like this.
 

Offline fivefish

  • Frequent Contributor
  • **
  • Posts: 440
  • Country: us
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #16 on: December 01, 2015, 08:24:47 pm »
One has to keep in mind there are no power transformers in these things... no galvanic isolation from mains.

If there's a catastrophic failure (say a short-circuit on the project you're working on, miswiring on your breadboard, etc... and the ATX failed dangerously... well you get the picture.)

Good luck.
 

Offline macboy

  • Super Contributor
  • ***
  • Posts: 2252
  • Country: ca
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #17 on: December 01, 2015, 09:58:56 pm »
One has to keep in mind there are no power transformers in these things... no galvanic isolation from mains.

If there's a catastrophic failure (say a short-circuit on the project you're working on, miswiring on your breadboard, etc... and the ATX failed dangerously... well you get the picture.)

Good luck.
Actually that is quite incorrect. These do use a transformer between input and output, in a forward converter topology. The ground on the secondary is connected to earth ground for safety purposes, but the secondary could be floated with a little modification (not that I condone that).
 

Offline Paul Price

  • Super Contributor
  • ***
  • Posts: 1419
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #18 on: December 01, 2015, 10:02:32 pm »
Fiveish:

Nonsense, ATX power supplies are as much galvanically  isolated as any other piece of equipment, such as a lab power supply or a desktop computer.

ATX power supplies do fail, but very rarely fail "dangerously." Mostly they just drop dead (no output) and so almost never damage the equipment they are supplying power to.
« Last Edit: December 01, 2015, 10:11:00 pm by Paul Price »
 

Offline Paul Price

  • Super Contributor
  • ***
  • Posts: 1419
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #19 on: December 01, 2015, 10:28:55 pm »
When my friend wanted a power supply to charge batteries and experiment with electronics, I converted an ATX power supply to a precision lab supply just for the challenge of it.
First of all, I was able to find and salvage as many ATX power supplies for free as I could pile up in my lab, so the cost of components to make a nice power supply was close to zero.
Taking one of them apart, I was able to salvage a large power toroid (about 3-in diameter) and used this to make a boost converter that transformed the +12V (>8amp rated) output to any voltage from a few volts(loaded) to around 30V. Then I used a LM317 mounted on a heat sink salvaged from one of the supplies to act as a linear regulator for a really smooth output.
I used a Microchip PIC18F25Kxx chip to control the output voltage of the boost converter I made with the power toroid that I salvaged.  The wire to wind upon the this large core already came free as removed from the junker ATX supply.
The PIC18 supplied the PWM drive to a 30-amp power MOSFET (free also from the motherboard of the same PC the power supply was taken out of) to drive the large toroid. The ATX supply conveniently supplied 12V for good gate drive for a power MOSFET.
The PICC also provided A2D to monitor and set the input and output voltage of the LM317 and also to drive 7-seg displays.
The same chip was used to monitor output voltage and current(With the help of a small valued sense resistor and an op-amp.)
So the same chip monitored the output current of the supply and allowed me to run the supply in constant current and constant voltage mode. Since the 18F chip had a built in 64-MHz clock, it was quite quick and able to adapt to change and provide 1024 PWM steps to adj. the voltage supplied to the LM317 and so I could carefully control the voltage difference across the LM317 to achieve maximum efficiency with just a medium sized heatsink. The fan to cool the supply was also free from an ATX, but was also controlled by the same PIC chip and was rarely on unless I was really drawing some amps from the supply.
« Last Edit: December 01, 2015, 10:38:15 pm by Paul Price »
 

Offline MarkF

  • Super Contributor
  • ***
  • Posts: 2523
  • Country: us
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #20 on: December 01, 2015, 10:42:14 pm »
Why not drop the ATX power supply and build a small power supply with a 3-terminal regulator.  Something fairly simple with a 1A current capability.  Check out circuit diagram.  Just add transformer, fuse and heat sink.
 

Offline Ludvig82Topic starter

  • Contributor
  • Posts: 18
  • Country: no
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #21 on: December 01, 2015, 11:12:52 pm »
Why not drop the ATX power supply and build a small power supply with a 3-terminal regulator.  Something fairly simple with a 1A current capability.  Check out circuit diagram.  Just add transformer, fuse and heat sink.

That looks like a nice project for later. I need some prefabricated boards for now. I dont have the soldering skill or equipment to start solder parts to pcbs yet. I can do some easy soldering like power to a board, and ofc wires. Hoping to get a nice soldering station with time :)
Never argue with an idiot, first he brings you down to his level. Then he beats you with experience!
 

Offline Ludvig82Topic starter

  • Contributor
  • Posts: 18
  • Country: no
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #22 on: December 01, 2015, 11:17:38 pm »
I think i got a shopping list put together now. Hopefully this will be a fun project :)

I got a PSU from an not so old computer.

Voltmeter to use on the variable output: http://www.ebay.com/itm/Mini-0-33V-Car-Motorcycle-DC-4-Digital-Voltmeter-Red-LED-Voltage-Panel-Meter-New-/390734584022
Buck converter: http://www.ebay.com/itm/DC-DC-Boost-Buck-Converter-Step-Up-Step-Down-Supply-Module-3-35V-to-2-2-30V-HG-/271720088700
A nice knob for adjusting the voltage (Will have to find a way to extend the pot on the Buck converter, and make it 6mm shaft.): http://www.ebay.com/itm/New-High-Quality-Control-Rotary-Knob-For-6mm-Knurled-Shaft-Potentiometer-/171474118517?hash=item27eca76775:g:dLoAAOSwd4tUI9vq&autorefresh=true

At the same time i will order some binding posts, alligator clamps and banana plugs :)

For an external case for this, i am not sure what i will do yet. Maybe i will go for wood.

I will post here when this projects starts :)
Never argue with an idiot, first he brings you down to his level. Then he beats you with experience!
 

Offline Paul Price

  • Super Contributor
  • ***
  • Posts: 1419
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #23 on: December 01, 2015, 11:17:49 pm »
MarkF: Good reasons not to just build a simple LM317 power supply are:
(1) You wouldn't learn much about electronics, or op-amps or current sensors or boost supplies or MCU programming, etc.
(2) You would have to pay for all the parts.
(3) You would need a very large heatsink and a fan.
      For instance, if you wanted to output 1.6 Amps at 1.6 Volts for an electrolysis experiment, the LM317 would be dissipating around      1.6A*24..say something about at least 40 watts of power. Makes your power supply a nice room space heater.
« Last Edit: December 01, 2015, 11:42:06 pm by Paul Price »
 

Offline Paul Price

  • Super Contributor
  • ***
  • Posts: 1419
Re: ATX Bench power supply, ideas and suggestions wanted
« Reply #24 on: December 01, 2015, 11:31:58 pm »
Ludvig82:

That boost module will not do the job. Whenever the input voltage to this device is greater than the output desired the output voltage is unregulated and tracks the input voltage.  So, if you supply the board with 12V you cannot set the output to less than approx 12V.

If you used the 3.3V supply of the ATX supply you could do a  little better and have a lower output voltage limit around 3.4V, but higher voltage output voltages around 20V would be not offering much current output.

Besides, the module you selected doesn't look very robust and would likely fail easily with a short circuit condition.

On the positive side, the knob choice is nice if it will fit the pot you haven't yet located, and the 4-digit LED voltmeter is just fine.
« Last Edit: December 01, 2015, 11:40:38 pm by Paul Price »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf